My VPS has 2gb RAM and CPU core is 1 and 1 TB HDD storage. The sync speed is very slow. Currently it has about 700 MB free RAM. bitcoind is taking about 66% of the 2GB ram. Should I add 1GB more RAM or is there any way to get similar improvement using any configuration?

3 Answers 3


This worked for me:

bitcoind -daemon -dbcache=4000
  • 1
    listen=1 is not required
    – user103136
    Commented Dec 15, 2021 at 18:50

First, make sure you have listen=1 in your bitcoin.conf file, and that the P2P port is open on your VPS so that it can make more than 8 connections.

Then you'll want to make sure you have a high disk speed or IO specified (definitely not HDD - I have a feeling you meant SSD). The faster the SSD the better. 2GB RAM should be fine, but you may want to up it to 4GB and 4 CPUs, at least while syncing. bitcoind will use the most resources while syncing, then you can downgrade once it's synced.

  • 3
    Adding connections does not speed anything up (generally you only sync from outgoing connections). Adding RAM is likely the most important thing (and equivalently increasing dbcache setting; otherwise more RAM doesn't help). Commented Mar 5, 2020 at 2:12

You can:

  • stop bitcoind; download the blocks from another source (there are some through torrent) and copy them to the data directory; You then restart bitcoind and let it go through the blockhain validating everything.
  • increase the database cache size with the command line -dbcache. You can increase this to max 4GB. But you'll need a more ram in your system.
  • 1
    More than 4GB works fine (only on 64-bit systems, of course), and is certainly beneficial. Commented Mar 5, 2020 at 2:15
  • -dbcache=2000 really helped here, although the OP may need to settle for 1200. Also temporarily stopping other process which were using RAM freed up more RAM for disk cache and dbcache. But dbcache was definitely the most helpful. When the sync is complete, I will remove -dbcache option because it's not needed. Commented Nov 3, 2021 at 21:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.